Dewey Benson Honeywell - NASA Benson Honeywell. ... MAU Module (Dual Slot I/O Card) Avionics have...
Transcript of Dewey Benson Honeywell - NASA Benson Honeywell. ... MAU Module (Dual Slot I/O Card) Avionics have...
NASA GRC
Propulsion Controls & Diagnostics Workshop
9 December 2009
Cleveland, Ohio
Dewey Benson
Honeywell
Honeywell Proprietary
Considerations for Distributed Engine Control
Key Technologies
Putting it all together – a system architecture
10
9
8
7
6
5
4
3
2
1 Air Valves (LPT) (HPT)
Turbine Case Cooling Valves
Air/Oil Cooler
Permanent Magnet Alternator / Generator
Air Turbine Starter
Start Bleed Valve
Turbine Vane Blade Cooling Air Valve
Cowl Anti-Ice
Fuel Metering Unit
Supplemental Control Unit10
9
8
7
6
5
4
3
2
1 Air Valves (LPT) (HPT)
Turbine Case Cooling Valves
Air/Oil Cooler
Permanent Magnet Alternator / Generator
Air Turbine Starter
Start Bleed Valve
Turbine Vane Blade Cooling Air Valve
Cowl Anti-Ice
Fuel Metering Unit
Supplemental Control Unit
Electronic Engine Control/FADEC
Health Monitoring Diagnostic System
Thrust Reverser Actuation System and
Shutoff Valves
Hydraulic Pump
Variable Bleed Actuator System
Inlet Guide Vane Actuation System
Fuel/Oil Heat Exchanger System
Lube and Scavenge Pump
Fuel Flow Metering Valve and Dividers
Surface Cooler
Nozzle Actuation
11
13
12
14
15
16
17
18
19
20
21
Electronic Engine Control/FADEC
Health Monitoring Diagnostic System
Thrust Reverser Actuation System and
Shutoff Valves
Hydraulic Pump
Variable Bleed Actuator System
Inlet Guide Vane Actuation System
Fuel/Oil Heat Exchanger System
Lube and Scavenge Pump
Fuel Flow Metering Valve and Dividers
Surface Cooler
Nozzle Actuation
11
13
12
14
15
16
17
18
19
20
21
Moving the electronics from single box here
To multiple nodes here
Presentation Outline
Lessons Learned• MAC modules successfully re-used across multiple applications• Network partitioning and composibility demonstrated• Great development schedule gains from systematic solutions to redundancy management• Dispatch Drives Fault-Tolerance => more fault tolerance is better.• Network hub required to assure system integrity
Hub is not conducive to physically distributed control system
Generation 1: Modular Aerospace Control ( MAC)
Distributed Engine Controls – Gen 1 Experience
Ma
in P
roc
es
sin
g C
ore
Po
we
r S
up
ply
Mo
du
le
Sp
ec
ial F
un
cti
on
I/O
Mo
du
leS
pe
cia
l F
un
cti
on
I/O
Mo
du
le
I/O
Mo
du
le #
2
I/O
Mo
du
le #
1
Po
we
r S
up
ply
Mo
du
le
Ma
in P
roc
es
sin
g C
ore
Lane A – Sensors and Actuators
HostProcessor
Bus InterfaceController
Power & I/OConditioning
Bus InterfaceController
HostProcessor
Bus InterfaceController
HostProcessor
Bus InterfaceController
HostProcessor
Power & I/OConditioning
Lane B – Sensors and Actuators
I/OConditioning
Bus InterfaceController
HostProcessor
I/OConditioning
Bus InterfaceController
HostProcessor
I/OConditioning
Bus InterfaceController
HostProcessor
I/O
Mo
du
le #
1
I/O
Mo
du
le #
2
Bus InterfaceController
Bus InterfaceController
Bus InterfaceController
HostProcessor
HostProcessor
HostProcessor
I/OConditioning
I/OConditioning
I/OConditioning
Hub A Hub B
• Born out of NASA funded research to
investigate modular certification
• Distributed, dual-lane configuration
• Based on TTP/C
• Smart sensor and actuation ‘nodes’
DevelopmentHigher cost for the first applicationLower cost for each new applicationFaster development and qualification time New systems would be composed of mostly existing modules
UpgradesFaster, cheaper upgrades – easily modify or add new funcitionalityObsolescence will cost less - only modules affected are redesigned
MaintenanceBetter fault isolationQuicker turn-around-timeLower maintenance costLower spares provisioning
Lower WeightDrastic reduction in wiring harness weight
Lower
Distributed Controls – Gen 1 Experience
Key TechnologiesNetwork Availability & Criticality Drives network and
system designHigh Temperature Electronics Cost vs QuantityDistributed computation Drives module designIntegrating smart sensors & actuators from different
manufacturers
Resources ArchitectureMore computational power is needed to support:
►Model-based Controls
►Predictive Health Management (PHM)
More sensors used to support PHMMore actuation needed to support advanced engine cycles
The Challenges
message message
message
node 1 node 2 node 3
The Integrity
AND
1
2
3
46
5
BRAIN Allows Integrity + Robustness Geographic brother’s keeper guardian action
Every node compares data forwarded from neighbor with data forwarded on skip link
Enables independence without separate silicon
Enables low complexity sensors and actuators
TT-GbE is Time Triggered Gigabit Ethernet 1 Gigabit copper or 10 Gigabit optical network
Deploying on NASA CEV program for Orion Vehicle
An aerospace-specific subset of Ethernet
TT-GbE is a scaleable, highly deterministic, fault-tolerant Ethernet
Compatible with avionics standard ARINC 664
Concurrent Availability + Integrity Without Hub
Key Technology – Networks
Distributed Controls - The Avionics Experience
MAU Module
(Dual Slot I/O Card)
Avionics have been integrating multiple suppliers’
equipment into a distributed system for years.
System Integration Through A ‘Virtual Backplane’TM
AT
A
Ch
ap
ter
Function
Biz
Je
t
RE
G (
ho
ld)
RE
G
21 Air Conditioning
21 Pressurization
24 Primary Power
24 Secondary Distribution
26 Fire Detection
26 Fire Suppression
27 Flight Controls
28 Fuel Gauging Embedded
28 Fuel Management
29 Hydraulics
30 Ice and Rain Protection
32 Anti-skid Braking
32 Nosewheel Steering
32 Landing Gear Control / Proximity System
33 Lighting
38 Water and Waste
49 APU Control
49 APU Door Control
52 Door Monitoring
74 Ignition
77 Engine Vibration Monitoring
78 Thrust Reversers
80 Starting
AT
A
Ch
ap
ter
Function
Biz
Je
t
RE
G (
ho
ld)
RE
G
21 Air Conditioning
21 Pressurization
24 Primary Power
24 Secondary Distribution
26 Fire Detection
26 Fire Suppression
27 Flight Controls
28 Fuel Gauging Embedded
28 Fuel Management
29 Hydraulics
30 Ice and Rain Protection
32 Anti-skid Braking
32 Nosewheel Steering
32 Landing Gear Control / Proximity System
33 Lighting
38 Water and Waste
49 APU Control
49 APU Door Control
52 Door Monitoring
74 Ignition
77 Engine Vibration Monitoring
78 Thrust Reversers
80 Starting
Module Host Software Host
Application provider / Integration participants- Smiths Industries- BF Goodrich- ELDEC- ABSC- Hydro-Aire- Messier-Dowty Electronics- Liebherr- Vibro-Meter- Hamilton-Sundstrand- Kidde- Whitaker- Kieser- Moog- Vickers- Parker- Multiple Honeywell Sites
The Avionics Experience - Integration of Multiple Suppliers
Distributed Controls –
Do Not Fear
9
Issues• High temp electronics have been the “show stopper” in past attempts:
• Lack of available parts to make a full distributed node
• Difficulty of getting non-volatile memory to work at high temp
• Cost of solutions
• User community wants the same cost as low temp electronics
• Compact size desired
Trade-offs• Standard silicon solutions versus Silicon-On-Insulator (or Sapphire) or SiC
• Build up solutions from discrete parts or develop custom chips (ASICs)
• Temperature capability versus reliability
Specific application needs could dictate different solutions
Key Technology – High Temp Electronics
2 Million Device Hours Worth Of Life Test Data
Standard Catalog Products:HTOP01 Dual Precision Op Amp HT1104 Quad Operational AmplifierHT1204 Quad Analog SwitchHTPLREG Voltage Regulators HT83C51 8-bit Micro ControllerHT6256 256Kbit SRAM (32K x 8)HT506 Analog Multiplexer (16:1)HT507 Analog Multiplexer (8:2)HTCCG Crystal Clock GeneratorHTNFET N-channel power FET
Custom Capabilities:• Gate Arrays
• MCM (Multi-Chip Modules)
• High Temperature Design Services
Products in Development:• HTA/D Converter (12 and 18 bit)
• HTEEPROM
• HTFPGA
• Reconfigurable Processor for Data Acquisition (RPDA)
High Temperature Part Availability Is Limited Compared to Bulk Si
Significant new parts developed from DOE funding
Key to flexibility:
Only High Temperature Non-volatile memory
Need collaboration to overcome high development costs
11
Take-awaysHow To Get The Most Out Of High Temperature Electronics
• Come at the solution from top and bottom
• Users need to define a cost point that still yields benefits
• Consensus design for high temp end products (node, data concen., etc)
• Develop accurate estimate around consensus designs
• Identify a flexible architecture – few chips cover all end products
• Incorporate into as few parts as possible to minimize cost Custom ASICs
• Users, suppliers, and gov’t share cost to develop high temp nodes
• Shared access to resulting products
• Use DOE Deep Trek program as model
First Attempt At High Temperature Electronics
• Tested at 200C
• Microcontroller w/ BIT
• 1553 Interface
• PWM Torque Motor Drive
• Position Sensor Interface
• 0.57 lbs
• 1.2 x 2.3 x 3.4 inches
Non-Volatile Memory problem solved
Hi-T Electronics Consortium being formed to solve cost
High Temperature Servoactuator Electronic Interface Unit (EIU)
Circa 1999, Funded by Air Force Research Labs
Detractors in 1999 were cost and inflexibility due to lack of non-volatile memory
C
CC
C CC
CC
CC
CCC
C
Airframe Switches.
IMA (front)
Generic Computational
Resources.
Networked via Gbs Time-
Triggerd Ethernet
IMA (rear)
Generic Computational
Resources.
Networked via Gbs Time-
Triggered Ethernet
Distributed On Engine
Components Networked
Via Fault-Tolerant
Ethernet 10mbs
BRAIN
10Mbs
Ethernet
BRAIN
10Mbs
Ethernet
BRAIN
Gbs -> 10mbs
Ethernet Interfacing Via
Normal Switching Action
Generic High
performance
Computational
resource
Cross-engine
Data Exchange
Via
Gbs Ethernet
• Shared computing resources
• Shared power supplies
• Distributed control
• Integrated Model-based controls
+ Vehicle Health Management
Putting It All Together – A Distributed Architecture
Most challenges are well understood and demonstrated: Distributed control
Smart nodes
Distributed computation
High integrity, high availability networks
What’s left? Define a common architecture
Define a common reusable, scalable set of parts
Move high temp electronics across the goal line!
Build and demonstrate an end system
Summary
Verification of Adaptive Control Techniques
Integrated Engine and Flight Controls
Engine Health Management Contribution to Flight Management Real-time flight assessment and fault accommodation
Continuous thrust available and fuel efficiency
Alternate flight plan solutions (i.e. return, divert, etc)
Combined Fault Accommodation of Flight/Engine ControlsControl reconfiguration to maximize flight envelope
Recognition of unstable configurations
Invoke extreme measures to regain stable configuration (e.g. overthrust)
Autonomous Civilian FlightSingle pilot cockpit No pilot capability
www.wondershare.com
Other Research Priorities